The DAM does this buffering and tracking of the different clocks on its own. In theory, any I2S signal gets "cleaned up" before any further processing occurs.
That is what is generally observed with DACs that have no FIFO buffers and no reclocking. The DAM is a little different - in theory (I keep saying that..) it should be immune to jittery I2S signals. But people have claimed that it is not 100% immune. I haven't really done any tests of my own. But I plan to this winter.. I've both a Kali and an Ian McFIFO buffer.
USB might sound like a bad idea, but if it's a proper implementation its output should have a lot less jitter than an RPi's / BBB's. But again, opinions vary.. in the end it's dependent on the specific implementation.
Thank you Dimdim so now I can build my system with Allo Isolator + Kali reclocker and DAM 1021. Before, I only see on the DAM 1121 have a pin strap to select Master or Slave mode.As you mention mean that DAM 1021 will automatic recognize I2S signal and change to Slave mode If have an external clock? If you try test Kali reclocker on R2R Dam soon, could you inform me about your comment
.The DAM1021 does not need and will not accept an external master clock. It must use its on-board programmable clock generator. The Kali will have its own clock domain, but that is not a problem because of all of the buffering and reclocking.
I haven't tested the Kali on the DAM, but if I remember correctly others have with no problems.
I haven't tested the Kali on the DAM, but if I remember correctly others have with no problems.
Ok, thanks again.
I see on your blog (very interesting by the way) that you have a Buffalo too - how do your find that compared to the DAM ?.
Ok. I intend to use the DAM to replace my Audiolab MDAC which also uses the ESS Sabre 9018. The Audiolab's USB implementation is rather poor (based on TI's TAS1020) with sonic resonance at high frequences).
Today i made some mods for my friend to soekris first revison of dac board, that you recomendet.
I did some measurements with 1khz and 10khz sinus wave. Can anybody tell me what is default filter settings for Soekris board...this looks like it is in NOS mode? He has less computer skills as I so he didn*t reprogramed any filter. Can*t be true that when you buy that board default setting is NOS... because that is not even close to specification for thd. I am not interested in NOS dac and i read that is R2R dac with proper measurement specifications. If you listen to that filter and describe how good sound is ... no comment.
Can anybody give me some hints how to change that ...i am not computer specialist.
Which filter should i choose or what do you recomend...i am not fan of a sharp filters.
Sorry for maybe hard words but i am very disapointed with measurement and sound is also ...
I try to do Tube output buffer for RAW output but lost my energy.
Thank you very much.
I did some measurements with 1khz and 10khz sinus wave. Can anybody tell me what is default filter settings for Soekris board...this looks like it is in NOS mode? He has less computer skills as I so he didn*t reprogramed any filter. Can*t be true that when you buy that board default setting is NOS... because that is not even close to specification for thd. I am not interested in NOS dac and i read that is R2R dac with proper measurement specifications. If you listen to that filter and describe how good sound is ... no comment.
Can anybody give me some hints how to change that ...i am not computer specialist.
Which filter should i choose or what do you recomend...i am not fan of a sharp filters.
Sorry for maybe hard words but i am very disapointed with measurement and sound is also ...
I try to do Tube output buffer for RAW output but lost my energy.
Thank you very much.
Attachments
AFAIK there has never been a "factory installed" NOS filter in any DAM.
Regarding the built-in filters, from the manual:
"FIR1, upsampling from incoming sample rate to 352/384 Ksps in one step, with different filter lenght based on incoming sample rate. All FIR1 filters are basic Parks-McClellan "brickwall" types, designed with http://t-filter.appspot.com/fir/index.html, but still shorter than your regular DAC.
IIR, bank of 15 biquads operating at 352/384 Ksps, with one used for the CD de-emphasis filter, none otherwise used for the basic DAC.
FIR2, upsampling from 352/384 Ksps to 2.8/3.1 Msps, reasonable short and soft but still using same design as FIR1.
All filters are using 32 bit coefficients, with up to 67 bit MAC accumulator."
Are you 100% sure that your board has the original filters set? Did your friend buy it direct from Soekris?
Did you try selecting one of the other available filters to see if the measurements change?
Regarding the built-in filters, from the manual:
"FIR1, upsampling from incoming sample rate to 352/384 Ksps in one step, with different filter lenght based on incoming sample rate. All FIR1 filters are basic Parks-McClellan "brickwall" types, designed with http://t-filter.appspot.com/fir/index.html, but still shorter than your regular DAC.
IIR, bank of 15 biquads operating at 352/384 Ksps, with one used for the CD de-emphasis filter, none otherwise used for the basic DAC.
FIR2, upsampling from 352/384 Ksps to 2.8/3.1 Msps, reasonable short and soft but still using same design as FIR1.
All filters are using 32 bit coefficients, with up to 67 bit MAC accumulator."
Are you 100% sure that your board has the original filters set? Did your friend buy it direct from Soekris?
Did you try selecting one of the other available filters to see if the measurements change?
Haven't measured my 3rd rev board, but it sounds very clean. Cleaner than my ciunas and ifi iDSD nano.
What power supply are you using? Transformer direct did not sound well to me. Simple lm317 regs work very well. I also use the un-buffered output.
Could there be a fault with one of your mods, or issues with the data stream feeding the soekris?
Sound from original to mooded is small but not huge improvement...
I don+t speak about sound now i just want to set dac to default filter...if this is default than i am done with that dac. I want to use it with RPI.
It is connected to regulated supply not direct...and i use Input board from DIY gruopbuy.
Dimdim can you do some measurement on your dac? 1khz and 10khz -10db sine wave as i did.
Thank you.
I don+t speak about sound now i just want to set dac to default filter...if this is default than i am done with that dac. I want to use it with RPI.
It is connected to regulated supply not direct...and i use Input board from DIY gruopbuy.
Dimdim can you do some measurement on your dac? 1khz and 10khz -10db sine wave as i did.
Thank you.
Last edited:
You will need a USB to serial port adapter. Have a look here, at the section "USB to Serie RS232 interface " : Soekris 's DAC implementations
Once you're in the serial console, you just type F4 or F5 or F6 or F7 to select one of the 4 available filters.
Stock filter should definitely not look like that...NOS on the other hand does. So either filters where swapped on that board or the software is defective, especially since there is no NOS in the original factory bank...
Hi, Soren,
Instead of developing a DAM to fit RASPBERRY Pi, why not developing a better or more hifi-dedicated single-board computer for your DAM? More interesting and maybe easier with all your knowledge and resources.
Cheers.
WYAN
Last edited:
Would be a wonderful thing to have available to DIY'ers.
But I suppose that the pricing could be an issue....
However, if particular measures are taken for audio, it may not have any real competition AFAIK, which could make it interesting business-wise....
For an SBC to become popular it has to:
1) Be cheap, so that
2) it gets supported by a large community of programmers & users.
That is the reason that the RPi is enjoying such widespread support.
If Soren (or anyone else FTM) were to design some SBC, he would have to either make it a closed project (and support it himself) or make it an open project (get its software started by himself and hope that the community takes it from there..).
In either case, it would be extremely time consuming for him and unless it was a closed project with an impressively big pricetag it would not be viable.
1) Be cheap, so that
2) it gets supported by a large community of programmers & users.
That is the reason that the RPi is enjoying such widespread support.
If Soren (or anyone else FTM) were to design some SBC, he would have to either make it a closed project (and support it himself) or make it an open project (get its software started by himself and hope that the community takes it from there..).
In either case, it would be extremely time consuming for him and unless it was a closed project with an impressively big pricetag it would not be viable.
Could it not be made to run on existing OS platforms, drivers, and players?
-or maybe I am missing something
But I agree that cheap/open or expensive/closed would be the typical business categories where audio-specific would lean towards the expensive and thus inhouse development of a certain magnitude...
The DAC, however, is in itself really great
-or maybe I am missing something
But I agree that cheap/open or expensive/closed would be the typical business categories where audio-specific would lean towards the expensive and thus inhouse development of a certain magnitude...
The DAC, however, is in itself really great
So you're talking about an 100% compatible RPi clone. That could work, but it wouldn't be much better than an RPi. If I were to have a go at it, I'd like to at least take care of the RPi's one major shortcoming when it comes to audio: its Ethernet port implementation. But that would brake RPi compatibility.
Plus I'm pretty sure Soren is confident that his DAC's FPGA takes care of all of that so from his standpoint all of this is pointless.
Plus I'm pretty sure Soren is confident that his DAC's FPGA takes care of all of that so from his standpoint all of this is pointless.